Does this target truly drive disease in humans?

Most drug development failures are not execution failures. They begin earlier, when a target was selected on evidence that correlated with disease without establishing that it caused it. Preclinical models cannot resolve this distinction. Human genetics can.

5 Prime Sciences applies Mendelian Randomization — a method that uses naturally occurring genetic variation as an instrument to estimate the causal effect of a target on disease outcomes in humans. Because genetic variants are assigned at conception, this approach is largely free from the confounding and reverse causation that undermine observational data.

Our analyses are powered by PRIMEVault™, a harmonised multi-omics database containing >70TB of genetic, proteomic, and metabolomic data — including 2,000+ GWAS, 15,000+ protein QTLs, and 1M+ eQTLs. This scale allows us to run proteome-wide and metabolite-wide analyses that go beyond single-target assessments, enabling portfolio-level prioritisation from a single body of evidence.

The output is a clear causal assessment: whether the target drives the disease of interest in humans, whether that effect is on incidence or progression, and whether the direction of modulation is aligned with therapeutic intent.

This evidence has directly informed a >$600M in-licensing decision and identified, ahead of the clinic, the causal gap that would have led to a >$100M Phase II failure.